The relation between inflammatory lesions and atrophy (global, tissue-specific, and regional) that is thought to represent neurodegeneration in multiple sclerosis (MS) is of fundamental importance in understanding the pathogenesis of this disease. We hypothesize that the activity and spatial location of the lesions drive the subsequent atrophy in MS. This hypothesis will be verified by analyzing the MRI data acquired on the CombiRx cohort. CombiRx is a multi-center, double blinded clinical trial with 1008 enrolled patients. Patients are being followed over a minimum period of 3 years with all patients followed until the last patient completes in January 2012 allowing for up to 6.5 years of follow up on some patients. MRI data on this cohort is acquired with a rigorous MRI protocol and the treatment assignments have remained constant. The five specific aims of this proposal are: 1) automatic identification of T2-hyperintense, T1- hypointense, and Gd enhancing lesions and their spatial location, 2) determine the cortical thinning that is a measure of cortical pathology,3) assess the pathology in the normal appearing brain tissue based on the T2 values, determined on a voxel-by voxel basis using the dual echo images, 4) determine the whole brain, tissue specific, and regional atrophy, and 5) determine the effect of lesion activity and their spatial location on the regional atrophy and examine the role of MRI measures as possible biomarkers/predictors of the disease. The image segmentation will be performed using the multi- spectral segmentation in combination with the atlas-based techniques. Activity of both T2-hyperintense and T1- hypointense lesions will be determined by subtracting images acquired at different time points following diffeomorphic nonlinear image registration. Regional atrophy will be determined using the tensor based morphometry. The effect of connectivity between the lesion location and regional atrophy will be investigated using the white matter atlas. Finally composite MRI measures will be correlated with both EDSS (extended disability status scale) and MSFC (MS functional scale) and their individual components. This strategy that includes spatial information should allow identification of robust biomarkers/predictors of the disease. The analysis based on a large and clinically well characterized cohort followed over a relatively long period of time to understand the relationship between inflammation and neurodegeneration is a unique feature of this proposal. Abbreviations: CNS (central nervous system); DGM (deep gray matter structures); DIR (double inversion recovery); DSI (Dice similarity index); DTI (diffusion tensor imaging); EDSS (extended disability status scale); FoE (field of expert); GM (gray matter); ICBM (international consortium for brain mapping); MNI (Montreal Neurologic Institute); MRF (Markov random field); MRI (magnetic resonance imaging); MRIAP (MRI Automated Processing); MRS (magnetic resonance spectroscopy); MS (multiple sclerosis); MSFC (MS functional composite); MTR (magnetization transfer ratio); NABT (normal appearing brain tissue); NAWM (normal appearing white matter); PD (proton density); RGM (regional GM); RRMS (relapsing remitting MS); RWM (regional white matter); SIENAX (Structural Image Evaluation, including Normalization, of Atrophy (X-sectional); SPM (statistical parametric mapping); TBM (tensor based morphometry); TOADS (Topology preserving Anatomical Segmentation); VBM (voxel based morphometry); WM (white matter)